The present invention relates to optical communication systems and more particularly to amplification in optical communication systems.
The explosion of communication services, ranging from video teleconferencing to electronic commerce, has spawned a new era of personal and business interactions. As evident in the rapid growth of Internet traffic, consumers and businesses have embraced broadband services, viewing them as a necessity. However, this enormous growth in traffic challenges the telecommunication industry to develop technology that will greatly expand the bandwidth limitations of existing communication systems. Further improvements in optical communications hold great promise to meet the continual demand for greater and greater bandwidth.
Wavelength division multiplexing (WDM) technology, in particular dense WDM (DWDM), permits the concurrent transmission of multiple channels over a common optical fiber. The advent of erbium-doped Fiber Amplifiers (EDFA) has accelerated the development of WDM systems by providing a cost-effective optical amplifier that is transparent to data rate and format. An EDFA amplifies all the wavelengths simultaneously, enabling the composite optical signals to travel large distances (e.g., 600 km) without regeneration.
One of the principal limitations of EDFA technology is limited bandwidth. Discrete and distributed Raman amplifiers have been developed to overcome this limitation. They provide very high gain across a wide range of wavelengths. Moreover, discrete and distributed Raman amplifiers increase the distance between optical regeneration points, while allowing closer channel spacing.
The operation of Raman amplifiers involves transmitting high-power laser pump energy down a fiber in a counter-propagating or co-propagating direction relative to the propagation direction of the WDM signal to be amplified. The pump energy amplifies the WDM signal.
One of the major limitations to the performance of Raman amplifiers (both discrete and distributed) is double Rayleigh backscattering of the signal resulting from amplification of certain unwanted signal reflections. It is known to ameliorate double Rayleigh backscattering by dividing up amplifiers into multiple isolated stages using different pumps, thus limiting the path length over which undesirable reflections may travel. This approach, however, leads to inefficient use of counter-propagating pump power, which cannot be readily distributed among isolated stages. Another approach relies on a complex configuration including 3 circulators and an interference filter and permits the energy of a single pump to be divided up over no more than 2 amplification stages.
What is needed are systems and methods for Raman amplification that ameliorate double Rayleigh backscattering while optimally employing pump resources and minimizing complexity. It would farther be desirable to allow the energy from a single pump to be distributed among as many amplifier stages as desired.